SUBJECT: Monitoring the centrifugal pump. 9-11

A sensible predictive maintenance program for centrifugal pumps is still an elusive dream for most plants. Unexpected pump shut down continues to be the rule. Most premature pump shutdowns are related to seal and bearing failures. The classic predictive maintenance techniques of machinery history, visual inspection and vibration analysis do not work well with products that fail prematurely, rather than wear out.

• Vibration analysis tells you that the destruction has already begun, not that it is going to happen. To use vibration analysis with mechanical seals we would have to know the frequency of a seal and with the many designs available, the wide choice of seal materials and the many fluids being sealed that is just about impossible.
• Machinery history is only valid if the equipment experiences traditional wear. Otherwise you are trying to predict an accident. Remember that no one wears out seals and bearings. They always fail prematurely. The L₁₀ life of pump bearings is rated in hundreds of years. Seals are supposed to run until the carbon wears down. When is the last time you saw one of those?
• Back up sealing is valid if you want to prevent unexpected seal shut down, but outside of purchasing a backup pump this approach is not practical for the bearings.
• Visual inspection of the remaining face carbon is possible with stationary seal designs by installing a pin that sticks out the back of the gland. This information would be valid if carbon faces wore out, but as you well know, it seldom happens.

If we elected to monitor the pump performance and use this data to predict an upcoming seal or bearing failure what exactly should we monitor? Lets look at some of the options:

THE WET END OF THE PUMP

You can monitor:

• The temperature at the pump suction and discharge flanges.
• The pressure at the pump suction and discharge flanges.
• A proximity gage can record the distance between the open impeller and the pump volute.
• Shaft deflection can be measured by proximity gages around the volute.
• Product flow can be measured by a variety of instruments without penetrating the piping.
• Strain gages could tell you if the rotating shaft has locked up when the pump is stopped.
• Vibration can be measured at several locations on the volute.
• Noise is easily detected and a valuable source of emerging problems.
• The amount of amperage being drawn by the motor combined with pump flow and capacity can be an excellent indication of pump performance.

THE STUFFING BOX AND SEAL AREA

You can monitor:

• Stuffing box temperature.
• Stuffing box pressure.
• Liquid leakage out of the stuffing box, or air leakage in.
• Stuffing box jacket inlet and outlet flow
• Stuffing box jacket inlet and outlet temperature.
• Seal gland flush pressure, flow and temperature.
• The temperature, pressure and flow of the fluid between dual seals.
• Convection tank temperature, pressure and level.
• Quench temperate and flow.
• Vibration.

THE BEARING CASE

You can monitor:

• Oil temperature to let you know if the oil is about to form varnish or coke.
• Oil level.
• Case pressure.
• Shaft movement or thrust
• The amount of water present in the oil.
• Shaft speed.
• Vibration
• Cooling coil inlet and outlet temperature, pressure and flow.

In an ideal preventative maintenance program all of these readouts would be incorporated into a single multi-pin outlet similar to the type found in all automobiles manufactured in the past few years. This outlet would then feed the information into a hand held computer that would be supplied with additional information retrieved off a bar code, on a tag, hung on the pump.

The bar tag information could be entered by anyone familiar with the process in addition to information supplied by both the pump and seal supplier. It would contain data about the fluid you were pumping, critical dimensions, and information about the bearing lubricant. It could include:

• The specific gravity of the fluid.
• The specific heat of the fluid and bearing lubricant.
• The temperature/pressure at which the fluids would change state and:
  • Vaporize.
  • Become viscous.
  • Solidify.
  • Coke.
  • Build a film on the seal faces or sliding metal parts.
  • Become a non-lubricant.
• The bar tag would also contain information about:
  • The correct impeller clearance to the back plate or volute.
  • The temperature limit of the bearing lubricant.
  • The maximum differential temperature across the pump.
  • The temperature limits of any of the seal components including the faces and rubber parts.
  • The pump BEP.
  • Desired flow through the cooling/heating jacket. You get this from the seal supplier.
  • Desired level, pressure and temperature in the dual seal convection tank.
  • The specified flush amount.

Now that we have an idea about what we can monitor, exactly what is it we would like to predict about pump performance? Wouldn't it be great to know the following?

• The pump differential pressure, flow and amperage tells us if the pump is running close to its best efficiency point? If it is of the BEP we're going to have lots of problems:
  • We will get excessive shaft deflection that translates to premature wear ring, seal and bearing failure. The impeller could deflect into the volute or back plate causing permanent damage to both pieces.
  • The lost power will convert to unwanted heat that can change critical shaft dimensions and tolerances. This can be a big problem in the bearing area where internal clearances are very critical.
  • You could break the shaft if the force generated is high enough.
• If the suction pressure decreases or the suction temperature increases there is a probability that we going to have a cavitation problem during the operation of the pump.
• Is the temperature or pressure change in the stuffing box going to affect any of the seal components? Many of these affects are non-reversible.
• There are many face combinations used in mechanical seals. Too many of them are sensitive to changes in temperature and pressure. Some ceramics, filled carbons and plated hard faces are especially sensitive to temperature changes.
• The elastomer (rubber part) is always sensitive to a temperature change either up or down.
• Corrosion always increases with an increase in temperature. This can be very important in acid applications.
• Seal flatness can be compromised in both high and low temperature applications.
• A temperature change in the stuffing box can tell us if the product going to change from a lubricating liquid to a non-lubricating gas or solid. Most of these changes occur when the pump is shut down or a cleaner or solvent is being flushed through the lines. Will shut down cause solid particles to appear in the fluid? Every fluid has a maximum and minimum operating temperature. Exceed these limits and all kinds of bad things happen. A change in stuffing box temperature or pressure can cause a lubricating liquid to :
  • Vaporize and blow open the lapped seal faces.
  • Crystallize and restrict the seal movement. Caustic is typical of this type of problem.
  • Become viscous and interfere with the seal movement.
  • Solidify between the lapped seal faces and destroy them, as well as restrict the free movement of the seal components..
  • Build a film on the sliding seal parts restricting their movement and separating the lapped faces. Both paint and hard water can do this.
  • Become a non lubricant. This is a problem with hot water applications that will lead to "slip stick" vibration problems between the lapped seal faces.
  • Cause the liquid to form solid particles that will get into the sliding components and restrict their movement. This is the "coking problem" we typically experience with all hot oil applications.
• Are the bearings in danger of failing?
• Is the lubricant temperature too high and increasing?
• Is the lubrication level too high or low?
• Has moisture penetrated the bearing case. Moisture can cause hydrogen embrittlement problems in the bearing.
• Are the seal faces glued together at start up? Any product that can solidify will cause this failure.
• When do you need an impeller adjustment? If you miss the clearance by as little as 0.002 inches (0.05 mm) you will lose one percent of the pump's capacity. This loss will be converted into heat and vibration.
• Do the wear rings need replacement? Internal recirculation wastes power and increases the pump internal temperature. Ten degrees centigrade (18°F) is considered the maximum temperature rise allowable across the pump volute.
• Are the seal's environmental controls functioning?
• Are you getting too much product dilution?
• Is the quench working?
• Is there enough stuffing box circulation to prevent the seal from being overheated?
• Is the cooling jacket becoming clogged from a build up in calcium?
• Is the inner seal of a dual seal application functioning?
• Is the flush fluid doing its job?
• Is the stuffing box being maintained at the correct temperature- especially at pump shutdown?

Now that we know what can be done, and any instrument technician should have no problem figuring out how to install the indicators, what are you going to do with the data you can collect? Here are some ideas.

First you need the base information:

• What should be the head, flow and power consumption at the best efficiency point? You get this information right off the pump curve. You will need the specific gravity of the fluid to convert the pressure reading from the gauge to head units so that you can read the pump curve. Be sure to adjust the numbers for the actual pump speed that you can read with a tachometer. Use the affinity laws for this.
• What is the maximum and minimum temperatures the product can tolerate without changing state from a liquid to a gas, crystal, solid, or becomes viscous ? Your facility knows more about this subject than any one else. Check with people in the engineering department or chemistry laboratory. Production people are another source of this information.
• What are the upper and lower temperature limits of the seal elastomer. The seal supplier can give you this information. Remember that there are different grades of various elastomers. Be sure you are getting the information about the grade you are using in your seal.
• Does the seal face combination have a temperature limit more restrictive than the elastomer? This is a consideration in most metal bellows seal designs. Make the seal people identify the material grade and have them supply the temperature limits.
• Some seal designs have restricted pressure or vacuum limits. Check with your seal supplier for this information. High pressure can cause elastomer extrusion and deformation of the lapped seal faces.
• What is the maximum pump inlet temperature or minimum suction head to prevent cavitation? The NPSHR information comes off the curve. Remember that the curve was generated using water as the pumping fluid. You will have to add the vapor pressure of your product to this number for an accurate NPSH required.
• What is the proper open impeller clearance? Get this from the pump supplier. You want the hot or operating clearance. You will need to use a cartridge seal if you are going to adjust an open impeller without interfering with the seal setting.
• What are the seal environmental control limits? The seal supplier has specified a pressure, temperature and flow in most cases.

Now that we have the base information and the pump readings, we should be able to prevent some of the most common seal and bearing premature failures.

• Is the pump about to cavitate? Cavitation can injure the seal components and shorten the bearing life. You must solve the problem before the cavitation begins. Cavitation can occur if :
  • The pump capacity increases.
  • The discharge head drops.
  • The suction temperature rises.
  • The suction pressure drops.
  • The outside diameter of the impeller is too close to the volute cutwater.
  • The speed of the pump increases.
  • Remember that the pump pumps the difference between the suction and discharge heads. If the suction head is increased and the discharge head is not increased the pump is now pumping at a lower head and the capacity will increase along with the possibility of cavitation.
• Is the product close to changing state in the stuffing box? If it does change from a liquid to a gas or solid the seal failure will soon follow.
  • Is the stuffing box temperature increasing?
    • Maybe the cooling jacket is not functioning. Calcium may be building up inside the jacket
    • Maybe there is too much flow through the cooling jacket. Remember that the cooling fluid should come into the bottom of the cooling jacket and out the top.
    • Has the shaft axial thrust over compressed the seal faces?
    • Was the stuffing box vented in a vertical installation?
    • Is there flow between the dual seals. Has convection stopped?
  • Is the stuffing box temperature decreasing?
    • Maybe the cooling jacket is too effective.
    • Is the buffer or barrier fluid between the dual seals at the correct temperature?
    • Is the stuffing box pressure dropping?
    • The discharge recirculation line may becoming clogged.
• Is the seal leaking?
• Have you accidentally hooked up suction recirculation instead?
• Is the impeller clearance correct? Too much slippage will generate excessive heat and vibration. This heat and vibration will translate to premature seal failure.
• Is the bearing oil too hot? If it is you're going to experience bearing failure.
  • Too high an oil level or overfilling with grease is the most likely problem.
• Face seals can maintain a positive pressure in the bearing case. As long as you have a positive pressure in the bearing case there is not much fear of water or solids penetrating into it. Water and solids along with the high heat caused by over lubrication are the main problems you have to prevent.

For information about my CD with over 600 Seal & Pump Subjects explained, click here  

 Link to Mc Nally home page